Like many other departments, Marketing has seen the benefits of data driven activities for the business. Online retail has revolutionized how we market to our customers, and while tech giants like Amazon and Google have made a business of marketing that drives customers to their sites or yours, most marketing departments aren't on the same level for driving business. However, with today's tools there is no reason your marketing department can't use data to see better results!
Before beginning to design a marketing analytics process (or any analytics process!), it is important to start with a strategy. What is most important for the marketing department to see first? Perhaps they need to understand how sales uses their marketing qualified leads better, or maybe they focus more on engagements. If they are happy going online to understand web activity but struggle with improving the success of email-based campaigns, embarking on web analytics first might not be the best way forward despite the relative ease of accessing data. Understanding the biggest pain points at the beginning will be invaluable when it comes to making your internal customers happy.
Once you've understood the pain points, the next step is setting up some Key Performance Indicators, or KPIs. Again, understanding what is important to the business and the marketing department in particular ensures that the analytics completed add value to the business. Establishing which KPIs to use before beginning any analytics ensures that everyone understands the current playing field and provides a quick reference to understand how marketing activities impact the business.
Common KPIs in marketing are impressions, leads, lead rates, engagements, engagement rates, wins, & win rates. Googling marketing KPIs will show you hundreds of other options, but that doesn't mean you should use all of them or any of the ones I listed! The important thing is that the KPIs are easily understood by those using them and that they relate directly to their everyday work. These could vary from team to team, and just because something isn't a KPI doesn't mean it can't still be provided in a report. I like to limit KPIs to 3-5, as not only does that prevent overload, but it also fits neatly across the top of standard dashboard size in Tableau (checkout Tableau's Distribute Evenly option if you struggle with neat KPIs)!
Once the strategy and KPI definitions have been established, it's time to go get the data you need. How you access and create reports can vary greatly, especially with so many marketing tools provided in-tool reporting options now. Depending on the size of the deparment and initial goals, these in-tool reports might be exactly what you need. If you have a larger marketing department or are looking to get a comprehensive overview across many marketing tools (and IT has not previously set up any data connections), it's time to start looking at APIs.
A variety of programs used by marketers such as Salesforce and Google Analytics provide easy access to their backend data. Tableau and Alteryx both offer standard connecters as well, with Tableau going so far as to provide data models for the most commonly used queries. While there are some limitations, these connectors offer the data to meet the majority of requirements for many basic marketing analytics projects. For programs such as Pardot, Eloqua, and Social Studio, both Alteryx and Tableau offer the ability to create custom connectors to APIs. Be sure to understand API limitations before embarking on these data extractions though! While all the programs listed above offer excellent documentation online, always leverage your system experts to be sure you pull the correct data.
Now that you've established the basics, you're on good ground to start providing some analytics to really drive your marketing department forward. There's an abundance of possibilities for marketing analytics, so let the list below help you get started!
One of the most common analytics approaches in marketing, this method of testing customer messages has been in use for decades now. A/B testing is easier than ever, as programs like Pardot & Social Studio offer built in solutions to meet this need that even allow users to understand the basic success of the test. Once you've set up data models with your data, diving into the results deeper is a quick and easy way to understand the best approach for your customers.
Sales teams aren't always willing to share the credit with marketing, with a common argument being "who's to say they wouldn't have bought our product anyways?" It's an age old marketing question, but that's no reason your department can't attach a dollar amount to the marketing results. Attribution models use metrics such as how many times a customer interacted with marketing materials or activities, previous purchases with the business, and the time between marketing interactions and purchases. These models can be heavily customized to each business, and lead to a solution that everyone in the business can be happy with.
How likely is your potential customer to actually buy your product? The more data the better to start with, but listen to sales expertise about what they think influences purchasing decisions to help direct you in choosing which factors to consider in this model. This is a more advanced analytics model, and python or R are typically required for these types of algorithms, but both Tableau and Alteryx enable you to easily use your code within the programs to quickly capitalize on actionable insights from the results.
Who are your customers and what have they bought from you?! An obvious question but not always one with an easy to find answer. This might require the purchase of some third party data to append onto your own customer information, but companies often find the insights invaluable and Alteryx makes integrating those datasets easier than ever.
Just how good are your leads? Are they really ready to hand off to sales? Creating a lead scoring model and leveraging programs like Pardot and Eloqua to implement them can help answer these questions, while providing a clear pool of leads for both sales hand off and marketing nurturing campaigns. If you want to ramp it up another notch, look into clustering analysis to uncover hidden segments of leads.